Alter Table in Parallel Job

[nohDS]

Is there a way to add/drop columns in a parallel datastage job during runtime?

For example: If I have a db2 table called Table1 with ColA and ColB, I want to be able to run a datastage job that can add ColC or remove ColA or ColB.

Is this possible?

[leandrohmvieira]

I think there is a way if there is some logic behind when drop and create columns, but IMO, this will be a little hard to maintain. Is there some special motive to drop and create columns on runtime?

Why not mantain ColA, ColB and ColC always on table and just fill it with nulls or dummy data?

[nohDS]

The columns ColA, ColB and ColC are only created depending on how many records there are in another dimension table. So if a customer were to stage more records into the dimension table, I would want a datastage job that can handle the addition of more columns.

[ray.wurlod]

You can use a Column Generator stage to add and drop columns.
You can use pretty much any stage to drop columns; Copy stage is probably the most efficient.
A dynamic approach might be possible using the Modify stage, or alternate streams of processing in the job design.

[chulett]

And, of course, your DataStage user would need to have the grants needed to alter tables... something that would never fly places where I have worked. Perhaps via an SP call, however.